Modelling driver's cognition is a challenging endeavour undertaken by psychologists and human factors researchers. The motivations underlying this enterprise range from safety concerns to improvement of traffic conditions. This approach is the next step after the production of risk models that explain drivers' behaviour. Before the advent of ITS, psychology applied to driving sought to understand and explain why accidents happen (risk management issues) and design countermeasures to prevent them. It turned out that many countermeasures did not meet the expected gain, so the appreciation of risks taken by the driver started to be explored by psychologists. This exploration led to another dimension of driving and the focus on "what does a driver understand of a given driving situation?" opened the door to the production of cognitive models. In this type of approach, the goal is not to explain why a driver takes a risk, but explain how a driver understands and processes a driving situation, propose how to assist drivers in their driving activity, and eventually design systems that will support them in their driving decisions and actions. The other well known challenge for these systems is to integrate the users characteristics, i.e. in order to improve the control of one driving task, the system should not degrade the other driving tasks that are carried out at the same time. This paper focuses on the design and implementation of the PADRIC (PATH DRIVer Cognitive) model and on the module in charge of reproducing part of the perceptive processing of the model. This model is integrated within a micro-simulation tool, SmartAHS, for supporting the development and assessment of driver assistance systems.
Samenvatting